Bayesian state estimation in the presence of slow-rate integrated measurement

Abstract

This paper concentrates on Bayesian state estimation approach in the presence of slow-rate integrated measurements. In chemical process, some quality variables, in the sense of measuring, often have an important characteristic resulting from the time taken for samples collection. These kinds of measurements are obtained based on sample of materials that are collected in a period of time. So, the measurement indicates the average property of the measurand in the period of samples collection, which is called Slow-Rate inTegrated Measurement (SRTM). In this paper, our goal is to estimate the fast-rate instantaneous states using available SRTM. Bayesian estimation approach is reformulated to acquire this goal. First, new Bayesian formulation is provided which ends to some complex integral formula. Then, with using the idea of Monte Carlo sampling method a numerical solution is presented for this problem. The advantage of the proposed algorithm is that it can deal with integrated measurement problem in a broader range of models with nonlinearity and non-Gaussian noise. The effectiveness of the proposed methodology is verified and demonstrated through simulation and empirical experiment on a level-flow laboratory-scale plant.

KEYWORDS:

• Slow-rate integrated measurement (SRTM)
• Bayesian state estimation
• sequential Monte Carlo
• soft sensor

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Fatemeh Yaghoobi

Fatemeh Yaghoobi received the B.Sc. degree in electrical engineering from Shiraz University of Technology, Iran, 2014, and the M.Sc. degree in electrical engineering from the K. N. Toosi University of Technology, Tehran, Iran, in 2017. She is currently a member of the Advance Process Automation and Control Research Group, KNTU. Her current research interests include industrial control systems, Bayesian state estimation, integrated-measurements, deep learning, and machine learning.

Alireza Fatehi

Alireza Fatehi received the B.Sc. degree in electrical engineering from the Isfahan University of Technology, Isfahan, Iran, in 1990, the M.Sc. degree in electrical engineering from Tehran University, Tehran, Iran, in 1995, and the Ph.D. degree in electrical engineering from Tohoku University, Sendai, Japan, in 2001. From 2013 to 2015, he was a Visiting Professor with the Department of Chemical and Materials Engineering, University of Alberta, Edmonton, AB, Canada. He is currently an Associate Professor of electrical engineering with the K. N. Toosi University of Technology, Tehran, where he is also the Director of the Advance Process Automation and Control Research Group and a member of the Industrial Control Center of Excellence. His current research interests include industrial control systems, slow-rate integrated measurement systems, machine learning, condition monitoring, nonmonotonic Lyapunov function, and advanced driver assistance systems.

Masoud Moghaddasi

Masoud Moghaddasi received the B.Sc. and M.Sc. degrees in electrical engineering from the K. N. Toosi University of Technology (KNTU), Tehran, Iran, in 2013 and 2016, respectively. He is currently a member of the Advance Process Automation and Control (APAC) Research Group, KNTU, Fault Detection and Identification (FDI) Laboratory, KNTU, and Single Unit Recording Laboratory, Neuroscience Research Center, Shahid Beheshti Medical University. His current research interests include machine learning, deep learning, computational neurosciences, intelligent systems and control, and applied statistics.